INTRODUCTION — The family of glycoproteins known as the hematopoietic growth factors (HGFs) plays a major role in the proliferation, differentiation, and survival of primitive hematopoietic stem and progenitor cells, as well as in functional activation of some mature cells. These effects are mediated by high affinity binding of the HGFs to specific receptors expressed on the surface of the target cells.
The major toxicities of the HGFs, the history of their identification, and an overview of their uses will be presented here. Their uses for specific clinical indications are presented separately. (See 'Clinical uses of hematopoietic growth factors' below.)
The function of specific HGFs in the development of blood cell lineages is discussed in separate topic reviews:
●Stem cell factor (see "Overview of hematopoietic stem cells")
●Erythropoietin (see "Regulation of erythropoiesis")
●Granulocyte and granulocyte-macrophage colony-stimulating factors (G-CSF and GM-CSF) (see "Regulation of myelopoiesis")
●Thrombopoietin (see "Megakaryocyte biology and the production of platelets")
HISTORY — Correction or amelioration of bone marrow failure by the administration of hematopoietic growth factors (HGFs) has been and continues to be a major practical goal of research in hematopoiesis. This goal could not be achieved, however, without the early tissue culture work, which led to characterization of the hematopoietic growth factor family, and without recombinant DNA technology, which provided the genes that allowed production of purified hormones in sufficient quantities to permit interpretable in vitro and in vivo studies.
Beginning with pioneering studies in the early 1960s, it has been recognized that normal and leukemic blood progenitor cells can be propagated in semisolid culture in the presence of soluble growth factors [1,2]. These factors were originally termed colony-stimulating factors (CSFs) because of their ability to support the formation of colonies of blood cells by bone marrow cells plated in semisolid medium [3].
During the 1970s and 1980s, it became clear that there were multiple types of CSFs based upon the different types of colonies that grew in the presence of the different factors. This observation led to the hypothesis that the growth and differentiation of blood cells were controlled, at least in part, by exposure of progenitor cells to CSFs having different lineage specificities [3,4].
Following the molecular cloning of the genes for many of these factors and their receptors during the 1980s and 1990s, it became possible to study in detail the structure, function, and biology of the recombinant CSFs as well as the molecular biology of their respective genes [3-6]. This analysis, along with similar work on the regulation of cells in the immune system, led to the realization that there is a large family of interacting regulatory molecules now generally known as cytokines or lymphohematopoietic cytokines, which control the hematopoietic and immune systems and integrate their responses with those of other systems [6-8].
This interacting network of cytokines includes the interferons [9], interleukins [10], tumor necrosis factors [11], and hematopoietic growth factors (including the colony-stimulating factors) [3]. The successful cloning of the HGFs and their receptors has provided an incredible array of tools for analysis of the molecular and cellular biology of hematopoiesis, and for the production of recombinant proteins for evaluation of the biology of the various factors in vivo and in vitro (figure 1).
IL-3 — The discovery, cloning, and expression of the gene for murine interleukin (IL)-3 (also called murine multi-CSF) presented the first opportunity to evaluate HGFs in an unambiguous fashion [12,13]. Sublethally irradiated mice were infused for seven days with recombinant IL-3 or control protein [14]. The spleens of the IL-3 treated mice were much larger than those of the controls, were more cellular, and contained more progenitors of the erythroid and myeloid lineages. In contrast, bone marrow cellularity was unaffected, although progenitor content was reduced. Similar results were obtained in mice injected intraperitoneally with purified bacterially synthesized IL-3 [15]. Other changes induced by IL-3 included 10-fold increases in blood eosinophils and two- to threefold increases in neutrophil and monocyte counts. Intraperitoneal injections resulted in six- to 15-fold increases in peritoneal macrophages with an increase in phagocytic activity.
These experiments clearly demonstrated that murine IL-3 influenced the replication and growth potential of primitive hematopoietic progenitors. They strongly suggested that the effects such hormones have on blood counts are related to their influences on progenitor function rather than peripheral blood cell kinetics. They also suggested that the function of mature cells can be altered in vivo, an effect that would be expected to decrease rather than increase the numbers of circulating phagocytes.
GM-CSF — The first indication that granulocyte-macrophage colony-stimulating factor (GM-CSF) also can broadly stimulate hematopoiesis in vivo resulted from studies in which GM-CSF produced in monkey kidney cells was infused into cynomolgus macaques [16]. Recombinant human GM-CSF (rhGM-CSF), when injected intravenously, has an overall initial half-time of 15 to 20 minutes, clearly demonstrating that infusion of the hormone at a concentration sufficient to maintain a functional blood level could be achieved. Infusions of rhGM-CSF into normal monkeys produced large increments in all classes of leukocytes, including eosinophils, lymphocytes, and reticulocytes [16]. The blood counts rapidly fell toward baseline when the infusion was terminated.
Mice that lack GM-CSF have normal basal hematopoiesis, but develop progressive accumulation of surfactant lipids and proteins in the alveolar space, the defining characteristic of idiopathic human pulmonary alveolar proteinosis [17,18]. (See "Causes, clinical manifestations, and diagnosis of pulmonary alveolar proteinosis in adults".)
G-CSF — In comparison to IL-3 and GM-CSF, granulocyte colony-stimulating factor (G-CSF) as well as erythropoietin and thrombopoietin are more lineage-specific. A primary effect of G-CSF is to promote the conversion of granulocyte colony-forming units (CFU-G) into polymorphonuclear leukocytes (figure 1). (See "Regulation of myelopoiesis".)
Mice that lack G-CSF have chronic neutropenia (20 percent to 30 percent of normal levels) and reduced bone marrow myeloid precursors and progenitors [19]. They also have a markedly impaired capacity to increase neutrophil and monocyte counts after infection with Listeria monocytogenes. Mice heterozygous for the null allele of G-CSF have intermediate values, suggesting a gene dose effect.
The efficacy of human G-CSF was initially evaluated in simian preclinical trials. Cynomolgus monkeys treated with two daily subcutaneous injections of purified G-CSF for 14 to 28 days showed a dose-related increase in polymorphonuclear neutrophils, with the plateau being reached after one week [20]. At the intermediate dose of 10 mcg/kg per day, total white blood cell counts of 40,000 to 50,000/microL were achieved, and neutrophil function was enhanced.
These initial studies also evaluated two cyclophosphamide-treated animals [20]. G-CSF was given either from 6 days before until 21 days after cyclophosphamide or for 14 days starting 3 days after the cessation of cyclophosphamide. The neutrophil count increased dramatically by day 6 to 7 after cyclophosphamide, reaching levels of 50,000/microL by the tenth day. On the other hand, the control animal remained pancytopenic for three to four weeks after treatment. This type of observation provided the rationale for the administration of G-CSF or GM-CSF to patients with chemotherapy-induced neutropenia. (See "Use of granulocyte colony stimulating factors in adult patients with chemotherapy-induced neutropenia and conditions other than acute leukemia, myelodysplastic syndrome, and hematopoietic cell transplantation".)
Erythropoietin — Erythropoietin (EPO) is essential for the terminal maturation of erythroid cells. Its major effect appears to be at the level of the CFU-E during adult erythropoiesis; recombinant preparations are as effective as the natural hormone [21,22]. EPO and its receptor may also contribute to wound healing responses, angiogenesis, and the response to brain and heart injury [23,24].
Recombinant human EPO has had a major impact on the quality of life in patients with marked anemia due to chronic renal failure as well as the anemia associated with cancer and its treatment. Its half-life in the circulation can be prolonged by the addition of N-linked carbohydrates (eg, darbepoetin) (figure 2), by formation of adducts with polyethylene glycol, or by preparation of EPO multimers [25]. (See "Treatment of anemia in nondialysis chronic kidney disease" and "Hyporesponse to erythropoiesis-stimulating agents (ESAs) in chronic kidney disease" and "Anemia of chronic disease/anemia of inflammation", section on 'ESAs' and "Role of erythropoiesis-stimulating agents in the treatment of anemia in patients with cancer".)
Thrombopoietin and thrombopoietin mimetics — Identification of the protooncogene MPL [26], based upon its homology to the oncogene transduced by the murine myeloproliferative leukemia virus [27], revealed an orphan HGF receptor that was important for megakaryocytopoiesis [28]. It also eventually led to the cloning of thrombopoietin (TPO), the ligand for this receptor [29-31].
Recombinant human TPO or its polyethylene glycol (PEG)-derivatized, truncated, 163 residue amino-terminus (PEG-megakaryocyte growth and development factor, or PEG-MGDF) stimulates megakaryocyte proliferation and endoreduplication in vitro and is a potent inducer of megakaryocytopoiesis and platelet production in vivo in mice and nonhuman primates [29-33]. The role of recombinant human TPO in the treatment of thrombocytopenia has been limited by antibody-mediated thrombocytopenia [34]. (See 'Clinical uses of hematopoietic growth factors' below.)
Eltrombopag and romiplostim are two small molecules that can activate the TPO receptor.
●Eltrombopag is an orally administered non-peptide drug that activates the TPO receptor by binding to its transmembrane domain. Eltrombopag given as a single dose to individuals without thrombocytopenia had no effect on platelet count, but daily doses for 10 days caused a dose-dependent increase in platelet count that peaked at 16 days [35].
●Romiplostim is an IgG1 Fc component linked to a peptide with no homology to TPO, which binds to the TPO receptor; it is administered by weekly subcutaneous injection. Single doses of romiplostim produce a dose-dependent rise in platelet count beginning on day 5 of administration and peaking at around two weeks [36].
Clinical use of these agents is discussed separately. (See "Clinical applications of thrombopoietic growth factors", section on 'Thrombopoietins (c-mpl ligands)' and "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists' and "Treatment of aplastic anemia in adults", section on 'EPAG alone'.)
Stem cell factor and Flt3 ligand — Stem cell factor (SCF), also known as Kit ligand or Steel factor [37], and Flt3 ligand [38,39], both interact with a variety of hematopoietic progenitor cells, perhaps most importantly with very early stem cell populations. While SCF has potent synergistic actions on early progenitor cells, its receptor Kit is also expressed on mast cells, and severe allergic reactions, including respiratory symptoms, have retarded clinical development.
CLINICAL USES OF HEMATOPOIETIC GROWTH FACTORS
Clinical settings — Several recombinant human growth factors (HGFs) have been evaluated in a variety of clinical settings. Largely because of availability, initial studies focused on GM-CSF and G-CSF in both transient and long-standing bone marrow failure syndromes, and on erythropoietin in the anemia of chronic renal failure. Following their successes, other HGFs such as macrophage-CSF, SCF, IL-1, and IL-11 have been evaluated, but none has found major clinical use.
The following is a list of the major clinical settings in which recombinant HGFs are administered. The efficacy of therapy in these conditions is discussed separately on the appropriate topic reviews:
●Transient bone marrow failure following chemotherapy (see "Use of granulocyte colony stimulating factors in adult patients with chemotherapy-induced neutropenia and conditions other than acute leukemia, myelodysplastic syndrome, and hematopoietic cell transplantation")
●Hematopoietic stem cell and progenitor cell mobilization (see "Sources of hematopoietic stem cells", section on 'PBPC mobilization')
●Recovery from hematopoietic cell transplantation (see "Hematopoietic support after hematopoietic cell transplantation", section on 'Growth factor support')
●Myelodysplastic syndromes (see "Management of the hematologic complications of myelodysplastic syndromes")
●Aplastic anemia (see "Treatment of aplastic anemia in adults", section on 'EPAG alone')
●Some forms of neutropenia (see "Overview of neutropenia in children and adolescents", section on 'Myeloid growth factors' and "Cyclic neutropenia", section on 'G-CSF' and "Congenital neutropenia", section on 'Treatment')
●Inherited bone marrow failure syndromes (see "Overview of neutropenia in children and adolescents", section on 'Myeloid growth factors' and "Overview of causes of anemia in children due to decreased red blood cell production", section on 'Diamond-Blackfan anemia' and "Shwachman-Diamond syndrome" and "Management and prognosis of Fanconi anemia", section on 'Transfusions and growth factors')
●Human immunodeficiency virus (HIV) infection-associated neutropenia
●Chronic anemias (eg, renal failure, prematurity, chronic disease/inflammation, HIV infection) (see "Treatment of anemia in nondialysis chronic kidney disease" and "Anemia of prematurity (AOP)" and "Anemia of chronic disease/anemia of inflammation" and "HIV-associated cytopenias", section on 'Indications for transfusion and erythropoietin')
●Preoperative erythropoietin to reduce the need for allogeneic red blood cell transfusion in other individuals who refuse blood transfusion (eg, Jehovah's Witnesses) (see "The approach to the patient who declines blood transfusion", section on 'Erythropoiesis-stimulating agents (ESAs/EPO)')
Thrombopoietin (TPO) was effective in increasing platelet counts in individuals with thrombocytopenia, but its use was associated with cases of severe autoimmune thrombocytopenia due to antibodies to TPO [40], and both TPO and a polyethylene glycol derivatized version (PEG-MGDF) were withdrawn from clinical use. Thrombopoietin receptor agonists were subsequently developed and are used in a variety of conditions (eg, immune thrombocytopenia [ITP]). (See "Clinical applications of thrombopoietic growth factors".)
Dosage and schedule — In adults, the recommended dose of G-CSF is 5 mcg/kg per day for most clinical situations other than peripheral blood progenitor cell mobilization, in which case a dose of 10 mcg/kg per day has been recommended [41]; infants and children with severe congenital neutropenia may respond to and be maintained on lower doses of G-CSF (approximately 3 mcg/kg/day). The recommended dose of GM-CSF is 250 mcg/m2 per day. Rounding the dose to the nearest vial size is an appropriate strategy to maximize cost benefit. The preferred route is by subcutaneous injection.
G-CSF is usually started no earlier than 24 hours after administration of chemotherapy [42]. Continuation until the absolute neutrophil count following the nadir exceeds 10,000/microL, as specified in the G-CSF package insert, is known to be safe and effective. However, a shorter duration that is sufficient to achieve clinically adequate neutrophil recovery is a reasonable alternative, considering issues of patient convenience and cost. G-CSF should not be given in the day or days prior to the next cycle of chemotherapy, or on the same day as chemotherapy or radiation therapy is administered. (See "Use of granulocyte colony stimulating factors in adult patients with chemotherapy-induced neutropenia and conditions other than acute leukemia, myelodysplastic syndrome, and hematopoietic cell transplantation".)
Dosing for erythropoietin depends on the indication. (See "Treatment of anemia in nondialysis chronic kidney disease" and "Anemia of prematurity (AOP)" and "Anemia of chronic disease/anemia of inflammation" and "HIV-associated cytopenias", section on 'Indications for transfusion and erythropoietin'.)
Dosing for thrombopoietic growth factors (eg, romiplostim, eltrombopag) is discussed separately. (See "Clinical applications of thrombopoietic growth factors" and "Immune thrombocytopenia (ITP) in children: Management of chronic disease", section on 'Thrombopoietin receptor agonists' and "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists' and "Treatment of aplastic anemia in adults", section on 'Treatments'.)
TOXICITY OF COLONY-STIMULATING FACTORS — Recombinant human GM-CSF and G-CSF have been tested in multiple clinical trials and have in general been well tolerated. However, a number of concerns have been noted (eg, transient leukopenia, systemic reactions, bone pain) [43,44]. There are more limited data on other growth factors such as M-CSF and stem cell factor (SCF).
Transient leukopenia — Following intravenous bolus injection, both GM-CSF and G-CSF induce a transient leukopenia in the first 30 minutes after administration. GM-CSF rapidly induces surface expression of the leukocyte adhesion protein CD11b (MO1) in vitro; expression of this protein is accompanied by an increase in neutrophil aggregation [45]. CD11a (LFA-1) and CD11c (gp 150, 95), two other members of this family of cell surface adhesion glycoproteins that have distinct alpha-chains but share a common beta-chain (CD18) with CD11b, are unaffected by GM-CSF. (See "Leukocyte-endothelial adhesion in the pathogenesis of inflammation".)
These findings have been confirmed by in vivo studies of sarcoma patients who received 32 or 64 mcg/kg per day of GM-CSF [46]. A marked increase of CD11b was noted that was evident by 30 minutes and persisted for 12 to 24 hours after treatment [46].
Radionuclide-labeled leukocytes are sequestered in the lungs after GM-CSF treatment [47], probably due to the aggregability and adhesiveness induced by increased CD11b expression. Breathlessness and hypoxia have been observed in some patients, particularly after short duration intravenous therapy.
In comparison to these findings, CD11b is not modulated by G-CSF. Thus, the mechanism responsible for the transient leukopenia following treatment with G-CSF is at present unclear.
Systemic reactions — GM-CSF can induce flu-like symptoms, including fever, flushing, malaise, myalgia, arthralgia, anorexia, and headache, and mild elevations of serum aminotransferases and rash are also reported. These effects are usually mild, are alleviated by antipyretics, and disappear with continued administration.
Pathogenic neutrophil infiltration (acute febrile neutrophilic dermatosis or Sweet syndrome) and cutaneous necrotizing vasculitis (leukocytoclastic vasculitis) can occur in some patients [48-52]. Upregulation of neutrophil function with the secondary release of cytokines may induce these complications. (See "Sweet syndrome (acute febrile neutrophilic dermatosis): Pathogenesis, clinical manifestations, and diagnosis".)
More serious systemic GM-CSF toxicity has been observed at higher dose levels (>32 mcg/kg per day intravenously or >15 mcg/kg per day subcutaneously). This includes a capillary leak syndrome manifested by weight gain due to fluid retention, pericardial or pleural effusions, ascites, and/or edema [53,54]. Phlebitis was noted in initial studies when GM-CSF was infused into small veins; large-vessel thrombosis has occurred with infusion of high doses into central veins [53]. No dose limiting toxicity has been observed with G-CSF.
Subcutaneously administered SCF frequently causes injection-site reactions [55]. It has also been associated with severe systemic allergic reactions that are thought to be mast cell-related, since SCF is known to activate mast cells [56].
Bone pain — GM-CSF and G-CSF both have been commonly associated with mild to moderate bone pain, coincident with or shortly after administration.
In a systematic review and meta-analysis of randomized trials comparing G-CSF with placebo or no treatment for the prevention of chemotherapy-induced febrile neutropenia, bone or musculoskeletal pain was reported in 20 percent of patients treated with G-CSF and 10 percent of controls (relative risk 4.0; 95% CI 2.2-7.5) [57]. A retrospective analysis of patient-level data from randomized trials comparing G-CSF versus pegylated G-CSF found a similar incidence of bone pain with both agents during the first four chemotherapy cycles (any bone pain: 66 versus 62 percent; grade 3/4 bone pain: 8 versus 7 percent) [58].
Bone pain is not limited to individuals with cancer or receiving cancer therapy. In studies evaluating G-CSF administration to healthy peripheral blood stem cell donors, bone pain was reported in 50 to 80 percent [59]. An assessment of bone pain from the National Marrow Donor Program reported bone pain in the majority of peripheral blood stem cell donors treated with G-CSF (95 percent); pain was mild, moderate, severe, and intolerable in 37, 48, 9, and 1 percent, respectively [60].
The etiology of CSF-associated bone pain is uncertain. Expansion of granulocyte progenitor cells in the bone marrow and elaboration of cytokines have been proposed as contributing factors [61]. Stimulation of hyperalgesia via CSF receptor engagement on neurons has also been suggested [61]. Occasional increases in leukocyte alkaline phosphatase and/or serum lactate dehydrogenase also have been noted.
Although data are limited, nonsteroidal anti-inflammatory drugs (NSAID) appear to be only modestly effective at reducing the frequency and severity of pegfilgrastim-induced bone pain. This was shown in a placebo-controlled trial in which 510 patients with a nonmyeloid cancer and no contraindication to use of an NSAID were randomly assigned to receive either naproxen (500 mg twice daily starting on the day of the pegfilgrastim injection and continuing for 5 to 8 days) or placebo [62]. Naproxen significantly reduced the overall incidence of bone pain (61 versus 71 percent) and its duration (1.92 versus 2.40 days). The incidence of severe pain (>5 on a scale of 1 to 10) was also significantly, but only modestly, reduced (19 versus 27 percent). Risk factors to predict incidence, severity, or ability to prevent pegfilgrastim-induced bone pain could not be identified.
Responses to antihistamines and opioids have been reported in those whose pain was not effectively treated with NSAIDs [61].
Antibodies to recombinant growth factors — rhGM-CSF that is produced in mammalian cells (Chinese hamster ovary [CHO] cells) is variably glycosylated on both O-linked and N-linked sites. In comparison, production in Escherichia coli results in nonglycosylated GM-CSF, while the yeast product is glycosylated only at N-linked sites. All three products appear to be equally effective, but antibodies have been reported in 4 of 13 patients given the yeast-derived product in phase I/II studies [63]. The IgG antibodies developed within seven days after the start of the infusion in all four patients, three of whom had received a bolus test dose. Antibodies were non-neutralizing as judged by bone marrow colony-forming assay, and were directed at sites on the protein backbone of the GM-CSF molecule that are normally protected by O-linked glycosylation, but which are exposed in the yeast and E. coli-derived products.
Antibodies to rhTPO have been reported in one patient with cancer [64] and in volunteers given PEG-MGDF. Further clinical development of the latter TPO formulation has been stopped, since transient decreases in platelet count were noted [33,65]. (See "Clinical applications of thrombopoietic growth factors".)
Possible stimulation of malignancy — Because HGF receptors are expressed by hematopoietic and several nonhematopoietic cell types, there has been a concern that certain malignant cell lineages might respond to such therapy, potentially worsening the underlying condition, or by triggering the development of malignancy in a susceptible individual [66-68].
An example of this concern has arisen in severe congenital neutropenia (SCN), the autosomal recessive form of which is Kostmann disease (mutation in the HAX1 gene). SCN may be associated with the acquisition of somatic clonal mutations in the G-CSF receptor. Affected children, if they survive infancy and early childhood, are at risk for myelodysplastic syndrome and acute myeloid leukemia; registry data suggest that the incidence of these complications in the G-CSF therapy era is 9.3 percent [68,69]. It has been speculated that G-CSF therapy might increase the risk of acute myeloid leukemia, perhaps more likely in those patients with G-CSF receptor mutations that can transduce a proliferative but not a differentiation signal [70]. At present, however, there is no evidence to support this hypothesis, as acute myeloid leukemia also occurs in patients who are untreated. It is more likely that the greatly improved survival associated with G-CSF treatment has allowed an underlying leukemia predisposition to manifest. (See "Congenital neutropenia", section on 'G-CSF receptor mutations'.)
Survivors of acquired aplastic anemia are also at increased risk for late myelodysplastic syndrome and acute myeloid leukemia. Development of these disorders has been associated with the acquisition of monosomy 7 and concerns have been raised about a possible contributory role from long-term G-CSF therapy [71-73]. In one patient, for example, the leukemic blasts were sensitive to G-CSF but not erythropoietin or IL-6 [72].
It is possible that long-term administration of G-CSF interacts with immunosuppressive therapy in patients with aplastic anemia. In one study of 72 adults with aplastic anemia, MDS developed in 1 of 47 patients treated without cyclosporine or antithymocyte globulin compared with 4 of 25 treated with one or the other of these agents [73]. All four had received long-term G-CSF at a higher cumulative dose than those who did not develop myelodysplastic syndrome; administration for more than one year was the single most important risk factor for myelodysplastic syndrome. Similar results were noted in another study in which MDS or acute leukemia developed in 11 of 50 children with acquired aplastic anemia who were treated with cyclosporine and G-CSF compared with none of 41 treated with either agent alone and none of 48 who underwent bone marrow transplantation [71]. (See "Treatment of aplastic anemia in adults".)
In order to deliver "dose-dense" chemotherapy regimens safely, G-CSF may need to be given during the short intervals between treatment courses. However, such an intensive schedule may increase the likelihood for the survival and proliferation of a hematopoietic stem cell that may have sustained a critical mutation from the previous chemotherapy course and would otherwise have undergone apoptosis or DNA repair.
Several observational studies reported that the use of CSFs is associated with an increased risk of therapy-related myeloid neoplasms (acute myeloid leukemia [AML] or myelodysplastic syndrome [MDS]). This issue was addressed in a systematic review of 25 randomized trials of chemotherapy with (n = 6058 patients) or without (n = 6746 patients) G-CSF for a variety of neoplasms [74]. AML/MDS was reported in significantly more patients treated with G-CSF (43 versus 22, relative risk 1.92, 95% CI 1.19-3.07). However, all-cause mortality was significantly lower in patients receiving chemotherapy with G-CSF support (absolute risk of death lowered by 3.4 percent, 95% CI 2.01-4.80), and greater reductions in mortality were observed in patients who received greater chemotherapy dose intensity.
Thus, the use of myeloid growth factors during chemotherapy increases the risk of a therapy-related myeloid neoplasm, although the absolute magnitude of the risk is small. The risk is probably outweighed by the benefits of using CSFs in this setting [44].
Possible enhancement of HIV replication — A concern with rhGM-CSF therapy, but not G-CSF, in patients with AIDS is the potential for stimulation of HIV replication. This phenomenon was initially demonstrated during in vitro experiments with mononuclear phagocytes exposed to rhGM-CSF or IL-3 [75]. Later in vitro studies revealed upregulation of CCR5 coreceptor expression and enhanced HIV infectivity in fresh human monocytes exposed to rhGM-CSF [76]. However, in vivo data on the relationship between rhGM-CSF therapy and HIV replication have been conflicting.
Multiorgan failure when used in sickle cell syndromes — Case reports have indicated that the use of G-CSF in patients with sickle cell syndromes (eg, homozygous sickle cell disease, SC disease, and S/ß+ thalassemia) has been associated with sickle cell crisis and multiorgan failure; at least one patient has died as a result of this complication. These agents are not used in individuals with sickle cell disease except in extremely rare circumstances (eg, life-threatening febrile neutropenia). (See "Overview of the management and prognosis of sickle cell disease", section on 'Avoidance of G-CSF'.)
TOXICITY OF ERYTHROPOIETIN — The largest experience with recombinant human erythropoietin (EPO) has come from the treatment of patients with anemia due to end-stage kidney disease. The most common side effects of EPO therapy, aside from hypertension and its related problems, are headache, which occurs in 15 percent of cases and an influenza-like syndrome affecting 5 percent [77,78]. The influenza-like syndrome is of unknown etiology, but is responsive to antiinflammatory drugs and does not seem to occur with subcutaneous EPO administration [79]. (See "Treatment of anemia in nondialysis chronic kidney disease" and "Hyporesponse to erythropoiesis-stimulating agents (ESAs) in chronic kidney disease".)
Hypertension, which is occasionally severe enough to be associated with encephalopathy and seizures [80], is the most important complication of EPO therapy [78,81,82]. Among patients with end-stage kidney disease, 20 to 50 percent of patients who receive EPO intravenously develop an elevation in diastolic pressure of 10 mmHg or more [78,82]. In comparison, the blood pressure is less likely to rise after subcutaneous therapy [83], possibly because this route of administration does not elevate plasma endothelin levels [84]. (See "Hypertension associated with erythropoiesis-stimulating agents (ESAs) in patients with chronic kidney disease".)
The risk of hypertension can be ameliorated by raising the hematocrit slowly [85] and by aiming for a hematocrit of 30 to 35 percent, a level that is sufficient to relieve symptoms without producing a significant elevation in blood pressure [81]. Patients who still become hypertensive can be treated with fluid removal (via dialysis or, if the patient has only chronic renal failure, diuretics) and the administration of antihypertensive agents. Beta-adrenergic blockers and vasodilators should be considered as agents of first choice, although calcium channel blockers and angiotensin converting enzyme inhibitors also may be effective. The dose of EPO should be reduced or discontinued for several weeks in severe cases or when other therapeutic measures are ineffective.
A number of cases of production of neutralizing antibodies to exogenous recombinant EPO, causing acquired pure red cell aplasia and refractory anemia in hemodialysis patients, have been described. (See "Pure red cell aplasia (PRCA) due to anti-erythropoiesis-stimulating agent antibodies".)
Potential harms of EPO when given to patients with malignancy are discussed separately. (See "Role of erythropoiesis-stimulating agents in the treatment of anemia in patients with cancer".)
FUTURE DIRECTIONS — Of all the hematopoietic growth factors (HGFs), two lineage-specific factors, namely G-CSF and EPO, have proven to be most useful clinically as stimulators of granulocytic and erythroid progenitors and precursors. They have been successfully employed in both transient and chronic bone marrow failure states, and, in a randomized trial employing EPO and supplemental iron, reducing the need for postoperative blood transfusion in patients undergoing total hip joint arthroplasty [86]. (See "Practical aspects of red blood cell transfusion in adults: Storage, processing, modifications, and infusion".)
Combinations of these late-acting factors with those that act synergistically on stem cells will certainly find roles in mobilization strategies for the harvesting of stem and progenitor cells [87]. (See "Sources of hematopoietic stem cells", section on 'PBPC mobilization'.)
A better understanding of the molecular mechanisms that control stem cell trafficking would potentially allow specific manipulation of these events. The importance of VLA-4, expressed on progenitor cells, and its receptor VCAM-1, expressed by stromal cells, in the localization of stem cells to the marrow may be important in this regard [88-90]. Antibodies to these molecules show that they play a role in the homing of stem cells to the marrow and in the mobilization of stem/progenitor cells from bone marrow to blood, an active mechanism that appears to require signaling through the stem cell factor (SCF) receptor Kit [91]. The stromal derived factor SDF-1 and its ligand CXCR4 on stem cells are important in the retention of HSC in the marrow. G-CSF administration leads to reduction of SDF-1 expression and to CXCR4 cleavage, thus mobilizing stem cells from the bone marrow to the blood; plerixafor inhibits the binding of SDF-1 to CXCR4 and can be used with G-CSF to mobilize HSC [92,93].
Another approach is the possible manipulation of self-renewal. As an example, murine bone marrow cells engineered to overexpress HOXB4 by retrovirus-mediated gene transfer show a dramatic upregulation in stem cell activity both in vitro and in vivo [94]. Serial transplantation studies have revealed a greatly enhanced ability of HOXB4-transduced bone marrow cells to regenerate the most primitive hematopoietic stem cell compartment, resulting in 50-fold higher numbers of transplantable totipotent hematopoietic stem cells in primary and secondary recipients, compared with serially passaged control cells. It is possible that overexpression of other transcription factors that drive hematopoietic development might allow stem cell division to be more successfully manipulated.
EPO mimetics — An additional area that holds great promise for future manipulation of the hematopoietic system is the development of small molecules, peptides, or agonistic antibodies that bind to specific receptors, with the objective of finding high affinity second generation drugs that may either be developed into oral agents [25,95] or generate molecules with longer-lasting activity [96].
●Oral agents can increase endogenous EPO production through interference with the oxygen-sensing pathway of EPO production [97,98]. Inhibition of prolyl hydroxylase stabilizes hypoxia inducible factor 1 alpha (HIF-1α), a factor important for EPO transcription; it also affects other iron homeostatic genes, lowering hepcidin and ferritin levels, which may obviate the need for concurrent iron therapy.
In a phase 2 study, the orally active prolyl hydroxylase inhibitor, FG-2216, increased plasma EPO in patients with renal failure and normal volunteers [99]. Three prolyl hydroxylase inhibitors, daprodustat (GSK-1278863), roxadustat (FG-4592), and vadadustat (AKB-6548) are being studied in patients with chronic renal disease. (See "Regulation of erythropoiesis", section on 'Hypoxia and EPO expression' and "Molecular pathogenesis of congenital erythrocytoses and polycythemia vera".)
●EPO mimetic peptides have been identified; their activity is enhanced by creating covalent dimers [100-105]. One of these agents, peginesatide (hematide), was approved by the US Food and Drug Administration for the treatment of anemia in adult patients with chronic kidney disease (CKD) who are receiving dialysis, but was withdrawn from the market due to hypersensitivity reactions [106].
●Continuous erythropoietin receptor activator (CERA) is a chemically synthesized "continuous" EPO receptor activator that is being tested in patients with renal failure and patients with anemia who are receiving chemotherapy [107].
Nonhematologic effects of EPO — Discovery that EPO may play a significant role in nonhematopoietic tissues (eg, brain, heart) has led to interest in EPO as a novel cytoprotective agent in both neuronal and vascular systems [108].
TPO mimetics — Small molecules that can activate the TPO receptor (MPL) are being studied for refractory immune thrombocytopenia (ITP), chemotherapy-and hepatitis C-associated thrombocytopenia, aplastic anemia, and myelodysplastic syndromes (MDS). Adverse events are mild (eg, headache, fatigue), but potential long-term adverse effects include bone marrow fibrosis/increased reticulin, thrombosis, increase in blasts in patients with MDS. These potential uses are discussed separately. (See "Second-line and subsequent therapies for immune thrombocytopenia (ITP) in adults", section on 'TPO receptor agonists' and "Clinical applications of thrombopoietic growth factors" and "Management of the hematologic complications of myelodysplastic syndromes", section on 'Thrombopoietin mimetics'.)
SUMMARY
●Recombinant hematopoietic growth factors (HGFs) are administered in the following clinical settings (see 'Clinical uses of hematopoietic growth factors' above):
•Transient bone marrow failure following chemotherapy
•Hematopoietic stem cell and progenitor cell mobilization
•Recovery from hematopoietic cell transplantation
•Myelodysplastic syndrome
•Aplastic anemia
•Some forms of neutropenia
•Inherited bone marrow failure syndromes
•Human immunodeficiency virus (HIV) infection-associated neutropenia
•Chronic anemias (eg, renal failure, prematurity, chronic disease/inflammation, HIV infection)
•Reducing the need for perioperative blood transfusion
●Potential toxicities of the recombinant HGFs include the following (see 'Toxicity of colony-stimulating factors' above and 'Toxicity of erythropoietin' above):
•Transient leukopenia
•Systemic reactions (eg, flu-like symptoms, capillary leak, hypertension, thrombosis)
•Production of deleterious neutralizing antibodies
•Possible stimulation of malignancy
•Possible enhancement of HIV replication
•Multiorgan failure when used in sickle cell syndromes
●The functions of specific HGFs (erythropoietin [EPO], granulocyte and granulocyte-macrophage colony-stimulating factors [G-CSF and GM-CSF], and thrombopoietin [TPO]) are discussed in detail separately. (See "Overview of hematopoietic stem cells" and "Regulation of erythropoiesis" and "Regulation of myelopoiesis" and "Megakaryocyte biology and the production of platelets".)
